The long-term goal is to understand how the brain learns and remembers. The purpose of learning is informed action: learned facts and events let us use the past to anticipate the outcome of familiar situations. Relevant memories must be available during learning if past events are to inform and be associated with ongoing experience. These abilities require the hippocampus and are devastated in Alzheimer's disease, but their neuronal coding mechanisms are unknown. In an "episodic-like" memory task, hippocampal activity varied with recent, ongoing, or imminent events when other aspects of behavior were identical, revealing current, prospective, or retrospective coding. The aim now is to identify firing patterns that predict learning. Aim 1 will assess hippocampal coding as rats learn new spatial behaviors in a familiar environment. Experiment 1 will oblige rats to make either new or familiar detours before completing a highly familiar, memory-based journey. Because the start and goal of the journeys are unchanged, the same coding of the familiar episodic-like memory structure should be maintained even if detours entail different trajectories. Journey-dependent coding should predict memory performance and rapid learning in novel detours. Experiment 2 will switch the start and goal of journeys in the same environment. Thus the rats will have to learn new journeys in the same spatial environment. Memory performance and journey-dependent coding should decline initially, and the acquisition of journey-dependent, coding should predict performance improvement during learning. Aim 2 will assess hippocampal coding as rats learn to perform the same memory task but in a new, unfamiliar environment, so the rats will have to learn new spatial representations as well as a new journeys. Journey-dependent and independent coding should be absent initially, but develop rapidly, and the acquisition of journey-related coding should predict performance during learning. Aim 3 will test if the identified dynamic coding properties appear even when learning is blocked. The same behavioral methods described in Aims 1 and 2 will be used, but one group of rats will be given NMDA receptor antagonists known to impair learning but not memory performance. The strong prediction is that neither stable memory performance nor accompanying hippocampal coding will be affected by the drug, whereas both learning and the dynamic coding properties that accompany it in Aims 1 and 2 will be impaired. Each aim will include hippocampus-independent tasks that require the same spatial behaviors but not memory, and will assess both population and temporal neural coding. The research should inform the design of rational treatments for amnesia (e.g. by stroke or Alzheimer's disease), including the future development of neural prostheses.